I am recording beautiful all marble, high ceiling spaces. THE REQUIREMENT IS 4 TRACK'S PER SAMPLING SESSSION. I am assuming that one will have to sample each location twice since Nat3 does 2 channels at the time.
The issue arises when phase aligning the two sample sessions. Giancarlo is it possible to supply me with a signal tone that has 3-4 very short clicks at the beginning to make phase alignmemt easier? One of the end products will be a 2D stereo reverb including location cues from four walls.
To recreate a reverb that reproduces information from all four walls ceiling and floor alike, one needs to be able to load two nebula instances over 4 channels in parallel. Phase alignmemt of the four channels is crucial. Are nebula3 plugin instances sample accurate in sync with each other?
I think you can't do that, because nebula always synchronize a stereo kernel, minimizing the phase offset if possible. So even if you sample it twice using an accurate testtone, those 2 nebulas will load the program syncronizing kernels by their own. You need a real 4 track nebula. Depending on requests we could implement that, but it would require time
Nebula reverbs have something that makes Nebula even more natural and interesting than impulse responses (by the way does Nebula use brute force convolution instead of Fourier Transform convolution). There is a real opportunity for Nebula to corner the reverb market by being able to include sound pressure, phase, HRTF information in a 4 channel format (WXYZ)this is how the human ear hears after all.
I will do some tests with the current technology and send you some programs to see if i can convince you of the benefit.
Giancarlo, isn't there an easier way to synchronise multiple nebula instances via some vst protocol, at least temporarily? Because with this technology there is more room for improovement by literally increasing the reverbs listener sweet spot using even more sampling channels (16-32). (I have research papers that support this if you are interested).
at the moment no, the engine tries to optimize samples. There are 4-5 synchronization rules (one even for hrtf impulses) but nothing for your case. If you want to experiment you could leave a loud dry signal in your path (I guess you have a big predelay, so it should work pretty well)